thank you very much Rick I’m very happy to be here thank you for the invitation it was a little bit intrigued by this topic because I’m not sure there’s much evidence that’s increasing in incidence but it gave me a few thoughts to start the talk with the first of which is what causes CLL in the first place what are the risk factors for developing CLL and so this is a summary from a recent study from the inter lymph consortium in which they did a pooled analysis of 13 case control studies 2400 CLL cases versus 15,000 controls and these were incident CLL cases and you’ll note that the top risk factor was a family history of he malignancy including any non-hodgkins lymphoma or leukemia other risk factors included having ever lived or worked on a farm having worked as a hairdresser having hepatitis C infection which is well known and protective factors included a history of atopic illnesses or sun exposure and so we don’t really have a lot the you’ll note that the odds ratios for these risks are not that high but in the case of the he malignancy family history this has been a very consistent risk factor across all studies and it’s probably underestimated in this study because it was not specifically asked of individuals it with incident CLL whether they had a family history if we look for example at Scandinavian databases where there’s good capture for both the pro bands and the relatives of hematologic malignancies we see that the risk factor for having a relative with CLL in an individual with CLL is actually 7.5 folds compared to the general population and risks are also increased for non-hodgkins lymphoma and hodgkin’s disease if we look in registry based studies such as for example ours the rate of reporting of first-degree relatives with the lymphoproliferative disorder among CLL patients is even higher on the order of twelve percent and perhaps about two-thirds of these report a relative with CLL i’ve noticed the fact that the population attributable risks namely the percentage of CLL thought to be due to inherited susceptibility is thirty-five percent for CLL which is quite high and the very low rate of CLL and East Asians is maintained amongst those who immigrated to the u.s. suggesting that this may be more genetic than environmental furthermore we see in the literature many sample pedigrees of families like this this is actually a family I’ll talk about later in our own database all of the affected individuals in the filled in circles have CLL and we see families like this with essentially McNeely inheritance not uncommon Lee and so in our registry that we’ve been running for 11 years you can see that the reported frequency of first-degree relatives amongst pro bands with non-hodgkin’s Hodgkins and CLL ranges from six percent with Hodgkin’s up to thirteen percent for CLL and if you include a second-degree relative they’re all above ten percent and it’s approaching twenty percent for the CLL patients amongst the CLL patients we see predominantly CLL but there’s also a significant rate of non-hodgkin’s lymphomas shown in blue there seems to be a little bit more crossing over to the other diseases amongst the CLL relatives than we see in the non Hodgkin’s and hodgkin’s Pro bands which is a fairly consistent finding and it’s likely that there is a phenotype for both just CLL and a family as well as CLL and other lymphoproliferative disorders and a family so what about this question of whether CLL is increasing so this is from the seer data which really doesn’t suggest that there’s been much increase maybe a bit of a blip in the 2000s and deaths are really quite stable I also queried the Mayo Clinic as to their opinion about this from their data and they agree that we’re not really seeing much incidents increased incidence of CLL or familial CLL but I think the question may have arisen because it does seem to me that at least in clinical practice I’m seeing more and more patients with so-called MBL or monoclonal be seldom Forsyte OSIS which was really only identity was identified earlier but it really only became more evident and codified about a decade ago and so these are individuals who have a monoclonal B cell population originally identified in residents near hazardous waste sites but also found in normal blood donors this led the subsequent studies of outpatients and clinic words found that about 2% would have potentially a CLL phenotype phone in their blood with no human illogic disorder and in those over 60 this number goes up to the five to six percent range depending on the sensitivity of the assay it actually can approach 75% ninety-year-old so this is quite common and unsurprisingly in unaffected members of CLL families or unaffected relatives of individuals with

CLL we also see more of this at the fourteen to eighteen percent rate with the relative risk highest in those who are younger because this is less common in the general population at younger ages like those less than forty and becomes more common with increasing age so the relative risk is less so it suggests that this could be a marker for the CLL risk phenotype and potentially a marker of having susceptibility to CLL amongst relatives of those with CLL and so this new diagnostic entity MBL monoclonal b-cell lymphomas was defined in 2005 as consisting of having circulating clonal b-cells most commonly with a CLL phenotype if it’s not a CLL phenotype but generally warrants additional work up but the absolute B cell count is less than five thousand and the patient’s have no other features of the lymphoproliferative disorder studies have suggested that all patients with CLL do go through a pre-existent MBL phase but the prevalence of MBL is several hundred fold more than CLL suggesting that most MBL cases don’t progress to CLL and so this led to a conundrum in 2005 because MBL was a b-cell count less than 5,000 but CLL was a lymphocyte count greater than 5000 but it turns out you can fairly commonly have a lymphocyte count greater than 5000 and a B cell count less than 5,000 this is actually true at that time of up to forty percent of newly diagnosed dry stage zero CLL patients so there was a potential overlap of the diagnostic criteria it turns out however that the B cell count of five thousand is a better threshold in terms of predicting treatment for your overall survival than the absolute lymphocyte count threshold and that’s why as you know the IWC ll in 2008 changed the diagnosis of CLL to requiring the absolute cell count of five thousand and so this led to further studies looking at MBL across populations and really the definition of two categories of MBL we have what we call population screening MBL people who have no evidence of enemy hematologic problem who are screened with highly sensitive flow cytometry and they typically have extremely low numbers of these clonal b-cells one per microliter on median versus those who are typically referred for an elevated lymphocyte count whose median count is about 2,900 and this is obviously much closer to CLL you can see that the distributions are fairly dichotomous and the official cutoff is 500 so it turns out this low count MBL has a very very low risk of progression to he malignancy there’s no difference in the rate of death or non he malignancy in these patients in a UK series and no effect on overall survival and perhaps unsurprisingly given that the biology of the cells differ amongst the low count MBL from CLL more frequently see by clonal or olive oil clonal populations that may come and go over time there’s a distinct igh repertoire that looks more like normal b-cells so they’re usually mutated and they don’t have the most common CLL V genes the 434 323 and 169 and stereotypy where an identical v gene is found across multiple CLL patients is very rare as our high-risk fish or cd38 or zap-70 but what about the clinical high count MBL so in multiple studies their risk of progression to treatment is about one to two percent per year shown here in data from the Mayo Clinic this is lower than for those with CLL B cells between five and ten thousand or greater than 10,000 and so differentiating this for your patients does have a potential prognostic impact in terms of outcome as one might expect the biology of clinical MBL is more similar to CLL but on the favorable lens compared to more advanced dry stages and these are data again from the Mayo Clinic looking at the survival of clinical M yell vs CLL in the upper left you can see that the dark line is the survival of clinical MBL which is not statistically different than the general population in that curve whereas it is for those with rise ero CLL with b-cell counts above five or ten in the upper right and lower left but in the lower right if you differentiate based on cd38 which was the primary prognostic marker they had they didn’t have adequate data on fish or igh those with the low risk cd38 are equivalent to the general population and survival whereas those with the higher risk are showing a worse outcome so this is just at the great majority the 80 or 90 percent of patients with clinical MBL who have lower risk markers do extremely well there are a couple things to watch out for though the incidence of second malignancies appears similar to CLL here are twelve to thirteen percent versus four percent in clinical controls and again here even controlling for treatment of CLL the MBL and CLL curves in blue are largely overlapping compared to the clinic population in red and

there’s also a higher risk of hospitalization for infection in high count MBL versus CLL again regardless of treatment for CLL and the right curve it’s sixteen to eighteen percent versus just three percent in the clinic controls here so as MBL increasing haven’t seen data to this effect but it hasn’t clearly been evaluated my impression from seeing patients is that this may be identified more and earlier as we’re looking at more CBC’s routinely and there may be a lower threshold for further evaluation and that particularly this may be done in those with a family history and so that we’re seeing these patients more and more who come in with MBL but a family history and I do think it’s important to differentiate for the sake of the patient from a prognostic standpoint patients are very happy to hear that this is associated with a better prognosis than the higher actual CLL and it’s also important to stress though that they need the same supportive care as CLL patients as we just saw in terms of screening for second malignancies vaccinations and care about infections okay so now I’m going to turn a little bit more back to familial CLL and first just comment on whether cases with a family history are different at all from those CLL cases not associated with family history traditionally we define familial CLS at least to first-degree family members in a family who have CLL and multiple studies have not found much difference in disease features between familial and sporadic disease in terms of stage a presentation need for therapy or overall survival even agent onset is not clearly very different and there’s not good evidence for anticipation which is when the disease occurs at a younger age in subsequent generations we looked at this in the crc database a number of years ago and found that fourteen percent of the patients had a family history of CLL amongst these referral centers and thirteen percent had a family history of another hematologic malignancy so that’s getting ups up to a quarter of the patients in this dataset only thirty percent of the patients had no history of any cancer it means age at diagnosis was similar there was a slight trend but only a year different and there were no differences in stage beta to igg cd23 cd38 or zap70 no difference in the frequency at which I ghv was mutated no difference in the selection of the v gene family and recently as part of our exome sequencing studies we deliberately enrich this population of 160 patients for familial CLL cases we had about thirty percent and we looked at the distribution of mutation rates amongst those with familial versus sporadic CLL and there was no difference similarly the frequency of the most common driver genes was not obviously significantly different between the familial and sporadic patients although a larger dataset would obviously be helpful in this regard so there’s really little evidence for biologic differences between familial and sporadic C lll those cytogenetics have not been adequately investigated in a couple papers in the literature including one of our own suggests the lower frequency of 11q deletion in familial CLL and we still need a large scale analysis of outcomes in comparison to sporadic CLL which is still hope to do through the CLL research consortium database now if we think about what are the models of genetic causation for a disease like CLL I think this is a very helpful diagram from a review by holsten’s group a number of years ago now if you think about it in relation to the number of cases in a family if there’s only one case that’s likely to be a sporadic event or at most there may be some low risk alleles in the family shown at the left of the slide and as more and more affected individuals occur in the family it becomes less likely to be sporadic low risk alleles are present but as you get up to four cases for example then it becomes almost Mendelian and it’s much more likely there’ll be a very high risk allele in that family and so if we just write that out a bit the high risk alleles would have a high relative risk of the disease but be very uncommon in the population and typically these would be identified by linkage and linkage has not been that fruitful in the case of CLL because it’s typically involved relatively small families with just two or three effectives included and not very large sample sizes there is one notable family that was reported by ohio state with a polymorphism in a regulatory region for the dap kinase gene in that family but not in any other CLL cases in contrast we have the alleles identified by genome-wide Association study shown at the bottom these are common in the population ten fifteen twenty percent of the population may have them but the relative risk conferred by each is very low often between 1.2 and 1.5 is on the high end of the risk associated with those genome-wide Association study alleles

still we now have five genome-wide Association studies that have identified up to 38 alleles probably 30 or so that are independent and they’re listed here most of them are not in jeans or coding regions they’re mostly intergenic and so the genes listed here are generally the closest gene I’ve highlighted in blue some of the ones that might mean something from biologic standpoint like irf4 milk I rf8 two of the more recent studies which has very large numbers thousands of cases identified a number of a pop tonic genes near g was– alleles we only have one case of a somatic driver that appears to be involved from a jiwa standpoint which was pot 1 10 of these have been tested as risk factors for MBL and in fact associations were observed for nine of them six of which were significant which is potentially consistent with the idea that MBL is a precursor condition for CLL now there’s a lot of interest in figuring out what these genome-wide Association study alleles do they’re highly statistically associated with the disease but they’re intergenic and so the hypothesis has been that their regulatory this was a study that found altered expression of two genes near g was– alleles this was in normal B cells we did a similar study ourselves in CLL cells where we found a strong association with expression of the hla dqa one gene shown at the top as well as FK RP at the bottom recently tried to extend this to be more mechanistic collaborating with Jay Brandner where he has been using an assay called a taxi which is a way of looking at chromatin accessibility with a transpose ace that cuts nucleus own free regions and we’ve looked at this in 15 c LLS and five normal b cells and compared regions of chromatin that look open or accessible by a taxi versus h3k27 which marks active enhancers and so in the upper left you can see that we find more of these in CLL than normal b cells or NBC and in the upper right you can see that most of the 74,000 such sites across the genome have been described in other diseases but 15,000 are unique to CLL about seventeen percent so they could be novel CLL regulatory regions and just to give you an example on the lower left we have the zap-70 gene we’re at the bottom in the NBC which is the normal b-cell you don’t see any significant spikes so there’s no expression of zap-70 which we would expect because that’s a t-cell gene above it we have a zap70 negative CLL which looks pretty similar to the normal b-cell not much expression at all but above that we have in red as AB 70 positive CLL where you can see there are some significant spikes in dick native of active transcription of the zap-70 gene and that’s statistically significant in the lower right so we use this to look at the 38 snips identified by G wasps in CLL and the top picture here looks at where those 38 genes are in terms of regions of the genome are they in coding regions or on translated regions in the top panels which you can see very few of them are are they intergenic or the bottom two rows are the h3k27 and the ATAC panels so on the top with just the G Wes alleles themselves we’re not seeing major involvement of the hypersensitivity regions we see about 12 of them overlapping with 38 snips but at the bottom panel if we look at all the G wasps alleles that are in linkage disequilibrium with or excuse me all of the other snips that are in linkage equilibrium with the jiwa snips in other words they could be the causative allele just as the jiwa snip because we can’t really distinguish them from the statistical studies that have been done there we see a very significant overlap with these regions of the genome that have it open enhancers and that’s illustrated in the middle panel here so this is the G wasps allele jiwa salil itself which doesn’t have open chromatin and this is the linked allele which is showing significant activity in the CLL versus the normal b-cell and here on the left this is the dish a random distribution of where we would expect these 38 snips to be if there were no selection for their association with the h3k27 or the ATAC and you see they’re black and black so they’re not significantly different but if we look at all links snips they’re very significantly enriched in these regions compared to a random distribution and much more so than any other cell type so this does suggest that many of these g West alleles maybe regulatory regions but what does this have to do with familial CLL so the frequency of these alleles and familial CLL has not been explored in detail we’ve looked at one

of them where we found the same frequency inspirado and familial CLL both of which were higher than the general population recent modeling suggests that these snips may explain up to fifty percent of the heritability of CLL but so far we’ve only explained nineteen percent and these analyses have had low power to delet detect low side with low risks or less common low sigh so there may be other types of alleles rare penetrant alleles that are similar to Mendelian alleles or less common snips with modest intermediate effects and so whole exome or genome sequencing can be used to find these it has the advantages of unbiased identification of all coding variation common and rare it’s enriched in rare variants but the statistics are challenging given the low frequency alleles and small effect sizes and not even all coding variants may have significant effects it’s very difficult to address non-coding variation as well this is a list from about a year ago of cases where next-generation sequencing is identified familial cancer genes and it’s mostly restricted to families thus far not populations and mostly Mendelian type inheritance so we’re doing this now co analyzing germline snips from 500 CLS and 9,000 normals focusing on rare variants and I think we’re almost there I’ve been saying that for a few years it’s a hard problem but I think we are getting there and then we’ll obviously need to evaluate candidates in terms of functional significance second hitting the tumors and concordance among family members another approach is to analyze family members 4shared novel variants but this is challenging because siblings share half their genome so I’m going to conclude by saying those are two families that we found that lymphoproliferative disorders are commonly familial the phenotype suvari shared genetic predisposition does not mean similar disease prognostic features linkage and association have not been helpful thus far families with Mendelian inheritance typically of a private allele common variation contributes somewhat but doesn’t explain heritability thus far and we don’t understand the biology of it but we’re starting to approach that with methods looking at regulatory regions source of remaining heritability is unknown but we hope to find it using exome and genome sequencing thank you

You Want To Have Your Favorite Car?

We have a big list of modern & classic cars in both used and new categories.